On 9/29/20 3:48 PM, Vladimir Sitnikov wrote:
> Andrew>You and I clearly have a different idea from what constitutes a
> concrete
> Andrew>proposal. This is hardly the ghost of a proposal.
>
> Can you please clarify what is a proposal from your point of view?
> Is it documented?
>
> I think I have read the relevant TODO items:
> https://wiki.postgresql.org/wiki/Developer_FAQ#What_do_I_do_after_choosing_an_item_to_work_on.3F
>
> Wiki clearly suggests posting a mail to pgsql-hackers before starting
> work.
>
>
A concrete proposal needs to be more than "a feature that does X". It
needs to contain a substantial implementation plan. What structures will
be affected, what APIS, what protocol changes and so on. You don't need
to have the code for these things but you do need a description of
what's intended by way of implementation that is detailed enough for the
community to critique. If you don't the danger is that you will spend a
lot of time coding and then present it to the community and they will
say "Oh, the design is all wrong." That's happened to a number of people
in the past, including some quite high profile people, and it's very sad
and frustrating for everyone when it happens.
A feature for streaming large data types could well be very valuable,
but right at the moment I at least don't have any idea what such a thing
could look like (and as you might imagine I'm quite interested).
cheers
andrew
--
Andrew Dunstan https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services